Welcome to our community

Be a part of something great, join today!

Log in
  • Hey all, just changed over the backend after 15 years I figured time to give it a bit of an update, its probably gonna be a bit weird for most of you and i am sure there is a few bugs to work out but it should kinda work the same as before... hopefully :)

Black shade temprature experiment.

Shaun Cammack said:
In the past, I had let the camera come up to temperature for 30 minutes and this still resulted in a cleaner image. Could be a fluke I guess. Seems reasonable, though.
Click to expand...

Hmmm, that's pretty interesting. What temperature did you blackshade at and what is your operating temperature?
 
Sidney L. Plaut said:
Hey Victor if you disagree with people experimenting calibrating at different temperatures , just don't do it. no need to argue that you disagree.

I have calibrated my Epic when its cold, medium and running hot, and there is a noticeable difference. And that, I believe, was the point of this thread.
Click to expand...

No useless arguing here. There's no need to do a lot of testing to know that there are differences.
 
The best thing to do for a blackshade is to trek up the highest mountain you can find and face into the wind, then come back down and shoot. If that's not practical, rent a boom lift and get as high as you can (facing the wind, of course). You can also call in the fire department and get them to raise you up if it's a really important shoot. I'd just call and explain to them the importance and I'm sure they'll send their best apparatus. If you can't manage any of that, the one secret trick is to spin around in circles with the camera upside down and have your camera assistant repeat 'Om Mani Padme Hum' until it's finished. The trick is that you really have to be in need to use this method, don't do it just as a shortcut or to save effort. I summited Everest with my Scarlet during a fierce storm and haven't had to re-calibrate since. The best results will come from the most extreme amounts of effort.

And one last thing to remember: don't ever, under any circumstances, for any reason or from any lapse of judgement, ever never ever blackshade at running temp. Don't do it.
 
Victor MOREIRA said:
No useless arguing here. There's no need to do a lot of testing to know that there are differences.
Click to expand...


Aesthetically - The only way to learn what suits me best at a given shooting situation, be it under cranking or hot/cold climate shooting - is through testing.... especially since the admitted and logical differences in sensor temperature, can result in images with less noise in the blacks.

If you have less noise in the blacks you could argue that the optimal dynamic range of the camera depends on a black shading done at a "correct temperature" at a given setting.... pretty important information to learn, share and discuss.
 
Victor MOREIRA said:
Yes
Click to expand...


I have been working all weekend. Will do this again on Monday. I think the reason it functions for me and my shooting is that i tend to be doing lots of long interviews lately, and the camera heats up. If im shading at the higher temperature, it makes perfect sense. perhaps for people who shoot short bursts, they may not see the benefit right away.
 
i get what you're saying timur, building in a small bit of temperature protection so you're always operating underneath your noise floor... pretty simple to understand. but some of this getting it as hot as you can get before emergency shutoff is absolute foolishness.
 
Timur,

I'm glad you started this thread, it's been a useful to me, at least. It explained some of what I was seeing and hadn't completely connected the dots that it's related to the black shade. I assumed that skipping black shading when moving between 'recording' temperatures was just going to be about more noise. The clips you shared showing the black shades at two different sensor temperatures surprised me--I wasn't expecting an exposure shift as well. Good information to have, perhaps I should have known that, but I didn't, so thank you for that.

Mark
 
QUOTE. "In the past, I had let the camera come up to temperature for 30 minutes and this still resulted in a cleaner image. Could be a fluke I guess. Seems reasonable, though."


With the current generation of cameras, sensor temperature management may always remain a dark art. By executing a "black shade" at a higher induced operating temperature, you may have "anticipated" a temperature condition the sensor would normally enter after a recommended warmup PLUS normal running time when temperature would have furthur increased, albeit maybe temporarily.

I imagine that if you were to draw a graph of the camera temperature from start up, there would be a steep rise which reaches a plateau after about 30 minutes. Run and idle cycles would show as small peaks and troughs with a heavy duty state causing a gradual rise of the plateau itself. So after reaching a near to normal temperature after a warm up period, your camera may be easing into the zone your black shade setting compensates for.


There could be several reasons why your method on the surface seems to yield the successful outcome for you. My pet theory on the main "why" your anticipatory method works is below.


I make some wild assumptions here so please treat the following comments with the same respect as bubblegum on your heel :-


By thermally "pre-stressing" the system, you are forcing it temporarily into its most common thermal state during an active duty cycle and sampling the black shade. Subsequently it enters into and does not depart from its black shade setting during each of the heat peaks.

The sensor, CPU and power management components will be the prime heat sources. I imagine with sensible design, these components will be thermally linked directly to the finned heatsink. I have not yet had opportunity to study closely the EPIC/SCARLETT casework. It seems that those major heat sources are actively cooled with a fan and finned heatsink in a ducted path through the camera body intended to isolate the other components from the heat generated by those parts. This is a good move as it reduces the mechanical loads on soldered joints and components more extreme temperature cycles would induce, hopefully extending their life.

Other miniature components on boards likely fend for themselves by passing their heat conductively to the board itself, radiation or by limited surface air contact. Final heat movement may be through passive vents in the casework or conductively through the case itself if it is sealed.

Unless the duct itself is thermally insulated downstream of the finned heatsink, it will itself become part of a closed loop by passing heat back to the heatsink via convective air movement and radiation within the whole enclosure.

This second cycle will be slower to establish itself and become stable. So in the thermal warm up graph in a fixed environment and no duty cycles, you may see an initial fast rise, a plateau with a long small slope and a final balanced position.


I think your experiment is a valid one. My personal preference based on your experience would be to save your "extreme" black shade" setting to call upon at any time and to do a "black shade" after a normal warmup time in a given environment.


Whilst the saved "black shade" settings can only be one custom recallable file plus the factory default setting, is there any scope for copying several custom black shade settings to a thumb drive and overwriting the in-camera custom black shade settings file from the thumb drive? My guess is that the saved settings may be in a write-protected zone, accessible only by the camera operating system and not the user. My thought was a good notion whilst it lasted.


The SI2K system has a write protection function which can be operator-disabled for firmware updates etc., then re-enabled. Among other things this permits one to customise the XP based human interface and some third-party applications like multichannel digital audio to the degree a normal computer will permit within sensible limits. One does this at one's peril but there is the recovery thumbdrive to fall back on if it all goes to hell due to operator error. My guess is that RED may have introduced a higher functioning donkeytrap which is enabled only by the firmware updates themselves to keep the uncontrollable and unpredictable third agency, the unsavvy human hand out of the mix as much as possible.
 
Blair S. Paulsen said:
Assuming that the ambient temps are moderate I agree that setting record and standby to the same rate can be less disruptive and depending on camera proximity might not even rile up the sound guy.
Click to expand...
If you can hear it, the microphone can hear it. If it doesn't bother the director, producer, or the post team, then it's not that big an issue. I will point out that there are other cameras that don't have a fan this loud. It can be a big problem if there's more than one camera in a relatively small, reflective room.
 
I would love a more technical explanation from Red on this. I'm aware of what's going on under the hood on a Phantom black shade but what I'm not aware of is why the Epic takes the amount of time that it does. What functions are being performed during the calibration process? I'm just curious, not being critical. Seeing as how the black calibration times have increased since the initial release of the Epic as there have been signifigant upgrades through the firmware,does that mean we should expect longer calibration times with the Dragon?



QUOTE=Les Dittert;1061798]The black shading on Epic takes so long because that part of the firmware is implemented in a 'turing machine' for true future proofing.
I also heard the black shading comes out better if the lens mount is facing Meca during the calibration.

This is all true, right Graeme ? ;)[/QUOTE]
 
Black shading is pretty much the only thing I dislike about shooting RED.

Was on a project that I needed to go from 24fps to a little bit of undercrank at 2fps. Had to sacrifice the blackshade and deal with with artifacts in post. Was paying $1000/day for the location and didn't have time to hang out for a blackshade.
 
Lee Saxon said:
I really hope they've addressed this nonsense with Dragon
Click to expand...

what nonsense? - a black shade is ofcourse needed to address core temperature vs noise in blacks... all digital cameras use black shades, some just do it faster... then again none have the specs of the EPIC nor the size.
 
I'm starting to think the forced higher temp black shade is akin to over-clocking a cpu in a computer. (I do believe I see better results--based on the cap-on, high-iso method--in the noise level.) This could explain why RED only recommends black shading at working temperature, to protect the sensor and cpu. To that end, I updated to firmware 3.3.12 and black shaded both cameras at working temps of 39/60. Once the seasons change I will black shade again at whatever the ambient temp calls for.

I'm pretty confident noise will not be a problem at these settings and if it is then I'll just apply a Neat Video fix.
 
I went a head a did a quick little test......and honestly I can not see a big difference.

Here's the info:
Black shading test
Camera: Scarlet/ Compression: 6:1 / Lens: Canon 16-35L focal length 35mm


C005 ISO 640 5600k f4.5 Normal User Blackshade (temp 33/42)


C006 ISO 640 5600k f4.5 Heated Blackshade (temp 40/60)


Light: Gekko Kelvin Tile (LED) camera left

You can download the R3D files here:
https://www.filemail.com/dl.aspx?id=LYLXWQXGHBWJNPN
 
Just wanted to note that I am getting a better result with a hotter black shade at around CPU 70c - I can now use 1600 ISO with very little noise. Most of what I shoot on Scarlet is unit (set) photography, which is often a hot environment. My camera regularly gets near 70c on set, so the recommendation to black shade at operating temp is a good one. However, I am also noticing an improvement at lower ISOs as long as the CPU is at least 50c. If you are having trouble with noise, it might be worth a try. Set those fans to manual (40-45%, depending on the ambient temp) and let the temperature rise!
 
The obvious guess is that the longer black shading times are RED's attempt to push the noise floor even farther down, particularly fixed pattern noise (FPN).

Whatever the specific provenance of the MX sensor, like virtually all such tech, it reflects a number of engineering compromises made during the design phase. Once placed in an actual device there are ways in which one can alter certain metrics like FPN rendition. My analogy would be Dolby noise reduction on audio cassette tape which reduced hiss with minor loss of high end content.

Bottom line, black calibration is intended to selectively neutralize low level nonlinearities specific to that particular sensor sample in actual operating conditions. High temperature black shading cuts noise while at the same time cutting sensitivity so its not a free lunch. OTOH, it does have the benefit of being based on actual evaluation of that particular sensor and can save you having to deal with noise in post. FWIW, on material that has more noise than I want, I "stub the toe" on the S-curve just enough to bury it in the blacks which effectively reduces sensitivity as well, just later in the process and with the visual feedback of a calibrated monitor in proper viewing environment to aid me. YMMV.

Cheers - #19
 
You must log in or register to reply here.
Back
Top